sdd-derive
Warn
Audited by Gen Agent Trust Hub on Apr 13, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSCREDENTIALS_UNSAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to read
.specs/.sdd/schema-config.yamland run "configured extraction commands" defined within (Phase 3). This allows for the execution of arbitrary shell commands based on content within the project's local directory. - [EXTERNAL_DOWNLOADS]: The instructions suggest the installation of an external CLI tool
code-review-graphusing the commanduv tool install code-review-graph(Phase 2). - [CREDENTIALS_UNSAFE]: The skill's codebase survey instructions (Phase 2) explicitly prompt the agent to find and read "Config or environment files," which commonly contain sensitive information like API keys or database credentials.
- [DATA_EXFILTRATION]: While no explicit exfiltration destination is identified, the skill reads sensitive configuration files and provides a mechanism to execute arbitrary commands, which could be leveraged to send sensitive data to an external server.
Audit Metadata